About DSLR cameras with video mode
In 2008 Nikon introduced the D90, the world's first DSLR camera capable of capturing High Definition video (HD). But why hadn't it been done before? Inside a DSLR there is a mirror which has to flip out of the way to expose an image, but as video demands a high number of frames per second this isn't mechanically possible.
Compact digital cameras, without an optical viewfinder, have no mirror - meaning capturing video has been possible for some time. However, until more recently they have been low resolution and of a low frame rate, usually 15fps. Movies, at least theoretically, need to be captured upward of 24fps to produce movement smooth to the eye.
To capture video, DSLRs take advantage of live view mode where the mirror locks up and a continuous live feed is displayed on the LCD. Simultaneously this data can then be captured.
An advantage of using a DSLR for video capture is that you can use your existing lenses, from wideangle through to telephoto. This gives greater creative flexibility, plus - due to larger sensor sizes found in DSLRs matching that of professional 35mm video cameras - the potential for a ‘Hollywood-style' shallow depth of field. Off-the-shelf video cameras tend to have small sensors which limits the depth of field.
So what's the catch and what are the limitations of using a stills camera to shoot video?
The concept of one ‘do it all' camera sounds great - just imagine swapping between stills and HD video capture at the flick of a switch. Well, now you can, with the Nikon D90, Canon EOS 5D Mk II, Pentax K7 and Panasonic Lumix GH1. Nikon and Canon also have entry-level versions in the form of the Nikon D5000 and Canon EOS 500D.
However, there are limitations to recording time. At full resolution, Nikon currently tops out at five minutes, Canon at 12 and Panasonic at 29. This is partly due to tax rates - a stills camera that can record more than 30 minutes of footage would, for tax purposes, be deemed a video camera. Other reasons include sensor overheating and buffer limitations as to how much data can be processed in any one sitting.
However, you can record sequences directly one after the other with a break in continuity. Although this may seem a hindrance, in reality it's rare to shoot a sequence for longer than a couple of minutes - look at any program on telly and you'll see most shots are held for a few seconds.